DOCSLIB.ORG

CAL Lisp: Preliminary Description

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
CAL Lisp: Preliminary Description ., CAL Lisp: Preliminary Description (Bill Bridge) Gene McDaniel Paul McJones A project for CS 220 and CS 199. / -1---' c> j . ,j.) '~;'4)':j ( ~.' . /1 , I . ! . ".--(-- ,,\--1.::> INTRODUCTION Lisp has been in the forefront of symbol manipulation programming languages for more than a decade and interactive implem~entations on timeshared computer systems have kept it up-to-date. In particular BBN Lisp for the SDS 940 and, more recently, the PDP-10 provide a very large address space and powerful primitive'; ) .. / operations allowing interactive debugging and editting ext~sions, among others, -t to be written in Lisp itself. (1,2). CAL Lisp is an attempt to coninue this line t of development and hopefully harness a fast computer, the CDC 6400, and a not-so­ r fast operating system, the CA~Timesharing system • ..; BBN has diverged from Lisp 1.S (3) in several ways which increase efficiency without, it is felt, detracting from the language. First, a linear pushdown stack replaces the association list scheme for variable bindings and allows compiled functions to reference their arguments directly. (These compiled functions still place the variable names with the values to facilitate debugging and maintain compatibility with interpreted functions.) Second, the property list of an atom is no longer used by Lisp to hold function definitions and constant values. Each atom has three settable attributes: a definition cell, referenced when the atom is used as a function name; a "zero-level value" cell, used when the atom occurs as a free variable bound nowhere in the stack; and a property list cell, purely for use by the user (including the Lisp-coded extensions) as a hatrack. Third, the treatment of special forms, functions taking unevaluated and/or a variable number of arguments, is nicely handled, with any of the four combinations allowed. A number of generalizations and extensions in many other areas exist and in PDP-10 Lisp provision is made for an openended set of datatypes, after the spirit of SNOBOL4. We wish to mimic the BBN Lisp faithfully enough to be able to utilize much of its library of software (editor, breakpoint routine, etc.) and yet still clean up a few of the inconsistencies which naturally occur as large systems evolve over the years. Our machine has a relatively large physical memory (23,000 words available to a user process) with an awkward 60 bit wordlength.... Secondary storage capacity is limited. with 8 million words of disk and 300 thousand words of extended core f ,: storage (~c~). Thus much of our effort has been applied to dense packing of data structures. And since no paging or segmentation hardware exists, we have had to carefully plan their inclusion in the software. The timesharing system does provide a disk file structure buffered by the ECS, and a software map to swap the central memory to/from files (residing on disk or in ECS.) The work is being done in BCPL, a high level language ostensibly suited for systems programming. This should greatly reduce the time spent on coding,but the resultant object program may be too large and slow to be feasible. We envision rewriting sections in assembly language as necessary. Th8 "\.-,orlo of' CJ\T, Lisp can be characterized by a complex of eleven con­ centric rings. F.ach ring represents a virtual machine, from the hardware (inner­ most ring) to the user program (outerm09t ring). The hardware consists of 32 K (K=102h) of eM (60 bit words), 1 CPU, 10 PPUs, 300K ECS and eight million words of disk storage. The ECS and disk act as secondary and tertiary storage devices. The lack of really large scale bulk storage imposes severe restraints on upon the upper layers of virtual machine. Next the ECS-INTERRUPT SYSTEM layer implements a number of orthogonal actions which aid in the implementation of a timesharing system. Of particular interest to Lisp is the ECS-CM interaction, ie., process swapping from and to CM and ECS. Associated with each user process is a software map which describes which blocks of CM are to be written where into ECS (and vice versa). The CAL Lisp system uses that software map to implement the virtual memory layer (described later). In the CAL Timesharing system files can exist either on the disk or in ECS. The low level disk system was designed to allow unused portions of a file to reside on the disk while the currently used portions of a file may reside in ECS. The directory system implements complicated searching procedures to allow the user to employ symbolic names while describing or asking for his files. The command interpreter-executor follows the two fold task of determining the "semantics" of user or subsystem requests and then attempting to service those requests. The virtual processor viewed from this point is the CAL Timesharing System. The lowest level processor directly implemented by CAL Lisp is the virtual memory mechanism. There is a 256kYirtual memory divided into lK of 256 word pages. The Lisp system references a virtual address through special macros: fetchvirtual(VA) and storevirtual(VA). They take a virtual address as input and return the current CM address 'of the word named by that virtual address. A page table is maintained which contains a one word entry for every 256 word page (See figure~). The page table, which is also paged, contains one word for every block in use, but need not contain entries for pages which aren't being used. A virtual address only need be right shifted a sufficient number of bits to produce the correct index in the page table. The virtual memory mechanism returns the CMaddress of the word being referenced if it is around; otherwise a call is made to the disk system to change the subprocess map. ! LISP PROCESSOR li AUX ,I eM AODR IC ::.1 ,U ~'s r(l'~ '''' c.a'(~ RF::.1 rU ~,s ~~~ ((ftre"ced ~,\\C~ lo.~t-liMe ~ pQ)'f! ~wllt, Io~ ~6: 1 if ~is i1 fA rud·OVl~J ..(il,r,'Y past LB:: tbfAtV ~ k ~~r 1'1'~ =- type '(14, b,-h ~:. du.y.illi",y ~rt ;,,~~J;on I i~J~J pse,",o·P'l)~ -if.sl AU~-::. b~h ~ 10 ~"41( 1k ,,~(ib~ dG.lt.~'pel j u.~ /U, . .f1... t c~" ~ \vSI etc. eM AUDIl ~ ~ck oJdrtn "{ I(~ 1 The disk system brings the missing file page into ECS from the disk and places it in the current subprocess map--thereby effecting the paged memory. This action may be preceded by the swapout-to-the-disk for some page chosen by the virtual memory mechanism. This occurs when CM is fully ma{ed with file blocks of the virtual address space (See figure 3 ). There is also a block buffer table which maps 256 word blocks of CM onto the virtual page currently residin~ there. Storing into a virtual address is more complicated: the paging mechanism must check of the virtual address being stored into belongs to a read only page. If so, a copy of this page is made, the r/o page is returned to the disk, ' the page table entry is changed to r/w and the appropriate CM address is returned to the caller. Important attributes of the virtual memory mechanism may be summarized in the following ways: 1) A large virtual address space is mapped into a small physical address space through use of software paging. 2) Software paging is accomplished by use of the subprocess map maintained by the operating system (which uses it to remove the user process to ECS when his time slice is up, or vice versa). 3) A Least Recently Used swapping strategy is implemented to decide which virtual page must be swapped out of CM to make room for a different virtual page. 4) When the system is smoothly running s1.apping experiments can be performed to determine good swapping strategies: the swapping mechanism may decide to keep more pages in ECS than there is room for in CM. This would mean that when a page fault occured it would not be necessary to retrieve the page from the disk. Virtual Segments The BBN Lisp partition~ the virtual address space into various types. An atom could only occur on a page within certain bounds in the virtual address space. While this allows for quick type checking by doing address arithmetiC, it also sets a static limit on the number of words which can be devoted to anyone type of object. CAL Lisp places no such address restrictions on objects. When a page is created, it is assigned to a specified type (eg., atom, string, etc). There is no limit to how many virtual pages can be devoted to a specified type beyond the intrinsic limit of the 256 K virtual memory_ Vir~l Mr~s This introduces the following problem: There is no .my of predicting in what order the pages of various types will be created. One page might be of type atom while the next two virtual pages might be strings or numbers. (See fiQ.1re 4 ). It is apparent that consecutive virtual addresses may lead to objects of completely different type. 1~ile all primitive functions should't have to worry about virtual address type, it is clear that chaos will result if virtual addresses aren't neatly "bundled." The virtual segment layer of the CAL Lisp system stands between the virtual memory and the data structures for the reasons described above. The virtual segment mechanism follows algorithms fitted specifically for the type of the data structure it may be handling at a given time.
Load more

Recommended publications
  • The Evolution of Lisp
    1 The Evolution of Lisp Guy L. Steele Jr. Richard P. Gabriel Thinking Machines Corporation Lucid, Inc. 245 First Street 707 Laurel Street Cambridge, Massachusetts 02142 Menlo Park, California 94025 Phone: (617) 234-2860 Phone: (415) 329-8400 FAX: (617) 243-4444 FAX: (415) 329-8480 E-mail: [email protected] E-mail: [email protected] Abstract Lisp is the world’s greatest programming language—or so its proponents think. The structure of Lisp makes it easy to extend the language or even to implement entirely new dialects without starting from scratch. Overall, the evolution of Lisp has been guided more by institutional rivalry, one-upsmanship, and the glee born of technical cleverness that is characteristic of the “hacker culture” than by sober assessments of technical requirements. Nevertheless this process has eventually produced both an industrial- strength programming language, messy but powerful, and a technically pure dialect, small but powerful, that is suitable for use by programming-language theoreticians. We pick up where McCarthy’s paper in the first HOPL conference left off. We trace the development chronologically from the era of the PDP-6, through the heyday of Interlisp and MacLisp, past the ascension and decline of special purpose Lisp machines, to the present era of standardization activities. We then examine the technical evolution of a few representative language features, including both some notable successes and some notable failures, that illuminate design issues that distinguish Lisp from other programming languages. We also discuss the use of Lisp as a laboratory for designing other programming languages. We conclude with some reflections on the forces that have driven the evolution of Lisp.
    [Show full text]
  • Lisp: Program Is Data
    LISP: PROGRAM IS DATA A HISTORICAL PERSPECTIVE ON MACLISP Jon L White Laboratory for Computer Science, M.I.T.* ABSTRACT For over 10 years, MACLISP has supported a variety of projects at M.I.T.'s Artificial Intelligence Laboratory, and the Laboratory for Computer Science (formerly Project MAC). During this time, there has been a continuing development of the MACLISP system, spurred in great measure by the needs of MACSYMAdevelopment. Herein are reported, in amosiac, historical style, the major features of the system. For each feature discussed, an attempt will be made to mention the year of initial development, andthe names of persons or projectsprimarily responsible for requiring, needing, or suggestingsuch features. INTRODUCTION In 1964,Greenblatt and others participated in thecheck-out phase of DigitalEquipment Corporation's new computer, the PDP-6. This machine had a number of innovative features that were thought to be ideal for the development of a list processing system, and thus it was very appropriate that thefirst working program actually run on thePDP-6 was anancestor of thecurrent MACLISP. This earlyLISP was patterned after the existing PDP-1 LISP (see reference l), and was produced by using the text editor and a mini-assembler on the PDP-1. That first PDP-6 finally found its way into M.I.T.'s ProjectMAC for use by theArtificial lntelligence group (the A.1. grouplater became the M.I.T. Artificial Intelligence Laboratory, and Project MAC became the Laboratory for Computer Science). By 1968, the PDP-6 wasrunning the Incompatible Time-sharing system, and was soon supplanted by the PDP-IO.Today, the KL-I 0, anadvanced version of thePDP-10, supports a variety of time sharing systems, most of which are capable of running a MACLISP.
    [Show full text]
  • Iliiiiiiiiiiiiwi1
    tt ¦ »:¦/ " : ^" : ¦ ¦ ' ¦¦ ¦ ¦ : WATERFORD STEAMSHIP COMPANY ¦ ¦ ¦ ' • : - ;; (LIMITED). "4^ ' ¦ - " C¦ rdrity arid :C ty i -[q f>i-: ^ ' t?^'fe w ¦¦;:|:!.r|: - Wdteiifor<M ^ ^ITA Mpi! , r!l ,: i !i) -|^i:U}, ^p; .!-J;. ' .:{ INTENDED ORDER OF SAILING. VISITORS to DUBLIN j QENERA.L QUABTEn : ; HLts)tTAlilA'f,ond ^i^M.: l &E$SI0FS} ' FEBRUARY, l<110 TRy THE ^TOECANiA"- j isJ'SiiD 0Nos of itHE d6cryirt,i are . the XlaisasS ' and , .Fti£teEA!hnx &els xn tiqOKfrfpRFHEy ¦ 1 Y8JU&f t9J0fL.;-\ ST&AiMIEKS • W«w'; t5t5ni«r i^FRANiOOiNIA;M ' FOUR ; ¦ ¦ ,i Swinr ¦ CLODAGH. DtlNBRODY. REGINALD, Byilqers' ironmo^g'ery. - ' -' : - ' " i::|' -THS ;. , ¦]¦• , ;; ; ; . i , 'i ' : LAKA ; ! ; : . : ': : ;,j v j ; ; ^ . Triple -sc&w, TuiMne | \\. HILAicVS; BB38IONS. -!,-|v • ¦V. , JTKNAPIA, Ett. COURTS 1 ¦ If OAEMANIA¦ ," Iiitnore.dfliB osincOT—Totsday <th January, 1910 • ' [ ' - . ,. : • CrowS ConTentent to- M ' -ifj 20,000. abni' J^T . : BmlDCsS—Tuesday, 4U) piouirj.'9 <? V t^md an4 "VTOTICE.—Tho O Tl PI Household Ironr^on^o ; BANIC Bralllr—Wedoesdiy,5th lanbary, ' '¦ ¦ Water* r^y- CITY :a C9UWTYiiLOAM €i|. blSGQUWT Twuwfecrow- Staapi«j "GARONIA-i'' 20,000 191a , ., .- : 'i : . ¦/ i "V— _3N. ii ford Steamship Cpra» \ : ¦ ¦ ' ' ' DutuRirrisV OTU Daaineov—T»unday,«lh . |ariua»y, Can't bo boaten (or Comfort ' «^€£"^m^ci3 1 ¦ 'a jpuUbei: of yjears) ' : ' : Bg 1910^Crosm BOSISBSS—Friday, 7U1 Jnnnaf^,1910 j Lour) , Pieces^PrbIO T" - (Wli' l ch la estAbHshed for '¦ ¦ ' !lV «iul j ChChi i mhey ' - " - ' ¦ ' ' ; ¦¦ ¦ ' 'i ^6r Stoairdcr i' Bprlty—Friday,^ ¦ ' ; nr) - 7th Jao;ary, ~S-raJ Pibces^^pgl ; - 'VCAlCPAiin-A. ' ^jSf*?l5b?''!! -I' " ima ' • < rt;ey . 1 •! ^WCS, iGowfi li . , ! ' I ' . \i . , - •!• •( ~ ¦ «nd Lite Stock foe Ship.
    [Show full text]
  • ' MACSYMA Users' Conference
    ' MACSYMA Users'Conference Held at University of California Berkeley, California July 27-29, 1977 I TECH LIBRARY KAFB, NY NASA CP-2012 Proceedings of the 1977 MACSYMA Users’ Conference Sponsored by Massachusetts Institute of Technology, University of California at Berkeley, NASA Langley Research Center and held at Berkeley, California July 27-29, 1977 Scientific and TechnicalInformation Office 1977 NATIONALAERONAUTICS AND SPACE ADMINISTRATION NA5A Washington, D.C. FOREWORD The technical programof the 1977 MACSPMA Users' Conference, held at Berkeley,California, from July 27 to July 29, 1977, consisted of the 45 contributedpapers reported in.this publicationand of a workshop.The work- shop was designed to promote an exchange of information between implementers and users of the MACSYMA computersystem and to help guide future developments. I The response to the call for papers has well exceeded the early estimates of the conference organizers; and the high quality and broad ra.ngeof topics of thepapers submitted has been most satisfying. A bibliography of papers concerned with the MACSYMA system is included at the endof this publication. We would like to thank the members of the programcommittee, the many referees, and the secretarial and technical staffs at the University of California at Berkeley and at the Laboratory for Computer Science, Massachusetts Instituteof Technology, for shepherding the many papersthrough the submission- to-publicationprocess. We are especiallyappreciative of theburden. carried by .V. Ellen Lewis of M. I. T. for serving as expert in document preparation from computer-readableto camera-ready copy for several papers. This conference originated as the result of an organizing session called by Joel Moses of M.I.T.
    [Show full text]
  • A Lisp Machine with Very Compact Programs
    Session 25 Hardware and Software for Artificial Intelligence A LISP MACHINE WITH VERY COMPACT PROGRAMS L. Peter Deutsch Xerox corporation, Palo Alto Research center (PARC) Palo Alto, California 94304 Abstract Data Types This paper presents a machine designed for LISP has many data types (e.g. list, compact representation and rapid execution symbolic atom, integer) but no declarations. of LISP programs. The machine language is a The usual implementation of languages with factor of 2 to 5 more compact than this property affixes a tag to each datum to S-expressions or conventional compiled code, indicate its type, in LISP, however, the , and the.compiler is extremely simple. The vast majority of data are pointers to lists encoding scheme is potentially applicable to or atoms, and it would be wasteful to leave data as well as program. The machine also room for a full word plus tag (the space provides for user-defined data structures. needed for an integer datum, for example) in every place where a datum can appear such as Introduction the CAR and CDR of list cells. Consequently, in BBN-LISP every datum is a Pew existing computers permit convenient or pointer; integers, strings, etc. are all efficient implementation of dynamic storage referenced indirectly. Storage is allocated in quanta, and each quantum holds data of allocation, recursive procedures, or only one type, so what type of object a operations on data whose type is represented given pointer references is just a function explicitly at run time rather than of the object's address, i.e. the pointer determined at compile time.
    [Show full text]
  • USENIX Winter Conference
    February ! 990 The Australian UNIX* systems User Group Newsletter Volume 11 Number 1 February 1990 CONTENTS AUUG General Information ..................... 4 Editorial ........................... 5 Secretary’s Letter ......................... 7 AUUG Inc 1990 Annual Elections -- Nomination Form ............. 8 Western Australian UNIX systems Group Information ............. 9 SESSPOOLE Information ...................... 10 AUUG Book Club Reviews ...................... 11 AUUG Book Club Order Form .................... 18 X Technical Bibliography ...................... 19 Commonly Asked X Questions ..................... 25 AUUG Institutional Members ..................... 42 From the EUUG Newsletter - Volume 9 Number 4 .............. 44 Editorial ......................... 45 How To Protect Your Software .................. 46 The EUUG Conference Mailing System ................ 54 UNIX in Czechoslovakia .................... 57 EUUG Munich Conference, Spring 1990 ................ 60 USING -- UNIX Systems Information Networking Group .......... 64 EUUG Executive Report .................... 67 Report From ICEUUG ..................... 69 News From The Netherlands ................... 71 UKUUG Report ....................... 74 USENIX Association News For EUUG Members ............. 76 USENIX Winter Conference ................... 78 EUnet Growing Up In Spain ................... 81 EUUG Software Distribution ................... 88 ISO/IEC JTC1/SC22AVG15 (POSIX) Meeting October, 1989 ......... 92 The OPEN LOOKTM Graphical User Interface .............. 98
    [Show full text]
  • 1 Introduction
    C7|Lisp Programming Russell Bradford 1 Introduction This course is about Lisp. Or rather this course is about how to program in Lisp. The distinction is a great one: Lisp, as a language to use, to program with, to learn, is one of the simplest of all languages. There is no difficult syntax, and every piece of code or data is treated in exactly the same way, with only very few (but important) exceptions. On the other hand, some of the hardest language semantic questions can be posed and written using Lisp, involving subtle or complex ideas, readily adapting to accommodate every new fad and paradigm. We shall concentrate on the programming aspect, and perhaps give the merest hint of the semantic under- pinning. 1.1 What is Lisp? The name \Lisp" means \List Processor." The basic object that we manipulate is the list. A list is an ordered collection of objects, each of which may be other lists, or things which are not lists, called \atoms." An atom is any object that does not have a list structure. Thus (the cat sat (on the) mat) is a list of 5 objects, the 4th being a list of 2 objects. The symbols the, cat, etc., are atoms. Here is another example of a list ((one 1) (two 2) (three 3)) which associates a symbol with a numeral (each of which are atoms). Lists come in all shapes and sizes, for example (((x 10) 2) ((x 5) ((y 2) 1) 1) ((x 2) 3) 2) is a more complicated list, which some people like to think of as a representation for the polynomial 2x10 + (y2 + 1)x5 + 3x2 + 2.
    [Show full text]
  • Ci the Interlisp Virtual Machine
    C.I THE INTERLISP VIRTUAL MACHINE: A STUDY OF ITS DESIGN AND ITS IMPLEMENTATION AS MULTILISP by OHANNES A.G.M. KOOMEN B.Sc. THE UNIVERSITY OF BRITISH COLUMBIA 1978 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Computer Science) We accept this thesis as conforming to the required standard. THE UNIVERSITY OF BRITISH COLUMBIA July, 1980 (c) Johannes A.G.M. Koomen, 1980 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Computer Science The University of British Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date July 22, 1980 ABSTRACT Abstract machine definitions have been recognized as convenient and powerful tools for enhancing software portability. One such machine, the Inter lisp Virtual Machine, is examined in this thesis. We present the Multilisp System as an implementation of the Virtual Machine and discuss some of the design criteria and difficulties encountered in mapping the Virtual Machine onto a particular environment. On the basis of our experience with Multilisp we indicate several weaknesses of the Virtual Machine which impair its adequacy as a basis for a portable Interlisp System.
    [Show full text]
  • Guide to the Herbert Stoyan Collection on LISP Programming, 2011
    Guide to the Herbert Stoyan collection on LISP programming Creator: Herbert Stoyan 1943- Dates: 1955-2001, bulk 1957-1990 Extent: 105 linear feet, 160 boxes Collection number: X5687.2010 Accession number: 102703236 Processed by: Paul McJones with the aid of fellow volunteers John Dobyns and Randall Neff, 2010 XML encoded by: Sara Chabino Lott, 2011 Abstract The Herbert Stoyan collection on LISP programming contains materials documenting the origins, evolution, and use of the LISP programming language and many of its applications in artificial intelligence. Stoyan collected these materials during his career as a researcher and professor, beginning in the former German Democratic Republic (Deutsche Demokratische Republik) and moving to the Federal Republic of Germany (Bundesrepublik Deutschland) in 1981. Types of material collected by Stoyan include memoranda, manuals, technical reports, published and unpublished papers, source program listings, computer media (tapes and flexible diskettes), promotional material, and correspondence. The collection also includes manuscripts of several books written by Stoyan and a small quantity of his personal papers. Administrative Information Access Restrictions The collection is open for research. Publication Rights The Computer History Museum (CHM) can only claim physical ownership of the collection. Users are responsible for satisfying any claims of the copyright holder. Permission to copy or publish any portion of the Computer History Museum’s collection must be given by the Computer History Museum. Preferred Citation [Identification of Item], [Date], Herbert Stoyan collection on LISP programming, Lot X5687.2010, Box [#], Folder [#], Computer History Museum Provenance The Herbert Stoyan collection on LISP programming was donated by Herbert Stoyan to the Computer History Museum in 2010.
    [Show full text]
  • Guide to the Herbert Stoyan Collection on LISP Programming
    http://oac.cdlib.org/findaid/ark:/13030/kt038nf156 No online items Guide to the Herbert Stoyan collection on LISP programming Paul McJones with the aid of fellow volunteers John Dobyns and Randall Neff Computer History Museum 1401 N. Shoreline Blvd. Mountain View, California 94043 Phone: (650) 810-1010 Email: [email protected] URL: http://www.computerhistory.org © 2011 Computer History Museum. All rights reserved. Guide to the Herbert Stoyan X5687.2010 1 collection on LISP programming Guide to the Herbert Stoyan collection on LISP programming Collection number: X5687.2010 Computer History Museum Processed by: Paul McJones with the aid of fellow volunteers John Dobyns and Randall Neff Date Completed: 2010 Encoded by: Sara Chabino Lott © 2011 Computer History Museum. All rights reserved. Descriptive Summary Title: Guide to the Herbert Stoyan collection on LISP programming Dates: 1955-2001 Bulk Dates: 1957-1990 Collection number: X5687.2010 Collector: Stoyan, Herbert Collection Size: 105 linear feet160 boxes Repository: Computer History Museum Mountain View, CA 94043 Abstract: The Herbert Stoyan collection on LISP programming contains materials documenting the origins, evolution, and use of the LISP programming language and many of its applications in artificial intelligence. Stoyan collected these materials during his career as a researcher and professor, beginning in the former German Democratic Republic (Deutsche Demokratische Republik) and moving to the Federal Republic of Germany (Bundesrepublik Deutschland) in 1981. Types of material collected by Stoyan include memoranda, manuals, technical reports, published and unpublished papers, source program listings, computer media (tapes and flexible diskettes), promotional material, and correspondence. The collection also includes manuscripts of several books written by Stoyan and a small quantity of his personal papers.
    [Show full text]
  • The Evolution of Lisp
    1 The Evolution of Lisp Guy L. Steele Jr. Richard P. Gabriel Thinking Machines Corporation Lucid, Inc. 245 First Street 707 Laurel Street Cambridge, Massachusetts 02142 Menlo Park, California 94025 Phone: (617) 234-2860 Phone: (415) 329-8400 FAX: (617) 243-4444 FAX: (415) 329-8480 E-mail: [email protected] E-mail: [email protected] Abstract Lisp is the world’s greatest programming language—or so its proponents think. The structure of Lisp makes it easy to extend the language or even to implement entirely new dialects without starting from scratch. Overall, the evolution of Lisp has been guided more by institutional rivalry, one-upsmanship, and the glee born of technical cleverness that is characteristic of the “hacker culture” than by sober assessments of technical requirements. Nevertheless this process has eventually produced both an industrial- strength programming language, messy but powerful, and a technically pure dialect, small but powerful, that is suitable for use by programming-language theoreticians. We pick up where McCarthy’s paper in the first HOPL conference left off. We trace the development chronologically from the era of the PDP-6, through the heyday of Interlisp and MacLisp, past the ascension and decline of special purpose Lisp machines, to the present era of standardization activities. We then examine the technical evolution of a few representative language features, including both some notable successes and some notable failures, that illuminate design issues that distinguish Lisp from other programming languages. We also discuss the use of Lisp as a laboratory for designing other programming languages. We conclude with some reflections on the forces that have driven the evolution of Lisp.
    [Show full text]
  • Stanford Artificial Tnteliigence Laboratory Memo AIM-228
    Stanford Artificial tnteliigence Laboratory July 1973 Memo AIM-228 Computer Science Department Report No. STAN-CS-74-409 The First Ten Years of Artificial Intelligence Research at Stanford Edited by Lester Earnest ARTIFICIAL INTELLIGENCE PROJECT John McCarthy, Principal Investigator HEURISTIC PROGRAMMING PROJECT Edward Feigenbaum and Joshua Lederberg, Co-principal Investigators Sponsored by ADVANCED RESEARCH PROJECTS AGENCY ARPA Order No. 457 COMPUTER SCIENCE DEPARTMENT Stanford University Stanford Artificial Intelligence Laboratory July 1973 tYlerno AIM-228 Computer Science Department Report No. STAN-CS-74-409 The First Ten Years of Artificial Intelligence Research at Stanford Edited by - Lester Earnest ARTIFICIAL INTELLIGENCE PROJECT John McCarthy, Principal Investigator HEURISTIC PROGRAMMING PROJECT Edward Feigenbaum and Joshua Lederberg, CoAprincipal Investigators ABSTRACT The first ten years of research in artificial intelligence and related fields at Stanford University have yielded significant results in computer vision and control of manipulators, speech recognition, heuristic programming, representation theory, mathematical theory of computation, and modeling of organic chemical processes. This report summarizes the accomplishments and provides bibliographies in each research area. This research runs supported by the Advanced Research Projects Agency of the Department of Defense under Contract SD-153. The views and conclusions contained in this document are those of the authors and should not be interpreted as necessarily representing
    [Show full text]